Departamento de Biologia Vegetal
INVESTIGATION OF THE GENETIC ETIOLOGY OF
SENSORINEURAL HEARING LOSS IN PORTUGUESE PATIENTS
ANA CLÁUDIA GASPARINHO GONÇALVES
Dissertação
Mestrado em Biologia Molecular e Genética
Departamento de Biologia Vegetal
INVESTIGATION OF THE GENETIC ETIOLOGY OF
SENSORINEURAL HEARING LOSS IN PORTUGUESE PATIENTS
ANA CLÁUDIA GASPARINHO GONÇALVES
Dissertação orientada por:
Professora Doutora Maria Helena Caria e
Professor Doutor Pedro João Neves e Silva
Dissertação
Mestrado em Biologia Molecular e Genética
Table of contents
Page
Acknowledgements... ii
Table of figures………. iii
List of abbreviations………. iv
Abstract………. vi
Resumo………. vii
Resumo alargado……….. viii
1. Introduction………... 1
1.1. Genetic Hearing Loss………. 1
1.2. DFNB1 locus……….. 1
1.2.1. GJB2 gene……… 2
1.2.2. GJB6 gene……….... 3
1.3. Low Frequency Sensorineural Hearing Loss.………... 4
1.4. Syndromic Hearing Loss……….……… 4
1.5. Mitochondrial Hearing Loss……… 5
2. Objectives……….... 7
3. Materials and methods………. 8
3.1. Sample characterization………... 8
3.2. Audiologic examination……….. 8
3.3. Extraction of genomic DNA from blood samples………... 8
3.4. Amplification by Polymerase Chain Reaction (PCR)………. 8
3.5. Screening of c.35delG mutation……….. 9
3.5.1. Restriction with BslI………... 9
3.6. Amplification of GJB2 coding exon………... 9
3.7. Screening for GJB6 deletions……….. 10
3.8. Mitochondrial DNA analysis………... 10
3.8.1. Detection of m.1555A>G mutation ………. 10
3.8.2. Detection of m.7445A>G and m.7511T>C mutations………. 10
3.9. Screening of WFS1 exons 4, 5, 6 and 8………... 11
3.10. Screening of SLC26A4 exon 6, exon 10, IVS8 and IVS14 regions……….. 11
4. Results and Discussion……… 12
4.1. GJB2 gene analysis………. 12
4.2. Screening of the c.35delG mutation……… 12
4.3. Automatic sequencing of the GJB2 coding exon……… 13
4.3.1. Case KQ………... 13 4.3.2. Family MV………... 13 4.3.3. Family NI………. 14 4.3.4. Case NP……… 15 4.3.5. Case NT……… 15 4.3.6. Case NY………... 15 4.3.7. Family OB……… 15 4.3.8. Family OF……… 17 4.3.9. Family OH……… 18 4.3.10. Family OI………. 18 4.3.11. Case OM………... 19 4.3.12. Family PD……… 20 4.4. Summing up on GJB2 screening………. 20 4.5. GJB6 analysis……….. 22
4.6. Mitochondrial DNA analysis………... 22
4.7. WFS1 analysis ……….... 23 4.7.1. Case BK……… 23 4.7.2. Family PF………. 25 4.8. SLC26A4 analysis………... 26 4.9. Pitfall………... 26 5. Conclusion……….. 28 6. References………... 30 7. Supplementary data ….………... 36
Acknowledgements
Um agradecimento especial a todos os que contribuíram para este trabalho:
À Professora Doutora Helena Caria, obrigada pela orientação, pela confiança que depositou em mim, pela transmissão de um espírito sempre positivo e confiante e pelos conselhos dados. Ao Professor Doutor Pedro Silva, obrigada pela orientação e ajuda dada ao longo deste ano. À Professora Doutora Graça Fialho, Coordenadora do Grupo de Surdez do BioFIG, obrigada pela oportunidade que me deu de integrar este grupo e por toda a ajuda ao longo deste ano. Aos pacientes, respectivas famílias e controlos intervenientes neste estudo, obrigada pela participação.
Ao Hospital de Egas Moniz, Hospital de Santa Maria e Centro Hospitalar de Coimbra, obrigada pela cedência das amostras.
À Helena Teixeira, Joana Chora e Tiago Matos. Os três foram importantes, cada um em etapas diferentes deste trabalho, mas todos me ajudaram e contribuíram para a minha aprendizagem ao longo desta etapa.
Aos colegas, do laboratório e não só. Egídia Azevedo, Nadieny Barbosa, Marta Delgado e Mónica Barra, obrigada pela ajuda, ensinamentos transmitidos e acima de tudo pelos bons momentos de riso que passámos. Um agradecimento à Professora Doutora Filomena Caeiro, pela sua preocupação e atenção constante.
Ao André Alcântara e ao Ricardo Cruz de Carvalho, obrigada pela ajuda que me deram com as questões informáticas, pela paciência e amizade.
À Manuela Lucas, obrigada pela amizade, pela ajuda, pela solicitude e pelos “puxões de orelhas” nos momentos certos!
À Sara Canato, obrigada pela sua amizade, pelos seus conselhos sempre em altura certa e sobretudo pela paciência em me ouvir!
Aos meus Avós e Pais, obrigada pela oportunidade que me deram de poder estudar, pela confiança demonstrada e por nunca terem condicionado as minhas escolhas.
Table of figures
Page Figure 1 - Schematic representation of the cochlea, showing the location of its different
structures……….. 2
Figure 2 - Map of the region on chromosome 13q11-12 showing GJB2, GJB6 and CRYL1 genes, relatively to centromere……… 3
Figure 3 - Schematic representation of pendrin protein………... 5
Figure 4 - The coding region of the 12S rRNA human mtDNA gene……… 6
Figure 5 - Schematic representation of human mtDNA showing the loci involving in mtDNA associated diseases………. 6
Figure 6 - Agarose gel from a c.35delG restriction analysis……….. 12
Figure 7 - Heredogram of Family MV………. 13
Figure 8 - Heredogram of Family NI………... 14
Figure 9 - Heredogram of Family OB……….………. 15
Figure 10 - Representative audiograms of Family OB……… 16
Figure 11 - Electrophoretograms of c.333-334delAA and p.Leu213X mutations, respectively……… 17
Figure 12 - Heredogram of Family OH………... 18
Figure 13 - Heredogram of Family OI………. 19
Figure 14 - Heredogram of Family PD……… 20
Figure 15 - Schematic representation of Cx26 protein……… 21
Figure 16 - Agarose gel from a GJB6 multiplex PCR………. 22
Figure 17 - Heredogram of Family NL……… 22
Figure 18 - Agarose gel from the restriction analysis of m.1555A>G mutation in mtDNA of Family NL……….. 23
Figure 19 - Pure-tone audiogram of individual BK1 and electrophoretogram of p.Asp171Asn mutation………. 24
Figure 20 - Heredogram of Family PF………. 25
Figure 21 - Schematic representation of wolframin protein……….. 25
List of abbreviations
12S rRNA – 12 Subunit ribosomal Ribonucleic Acid
ºC – degree Celsius δ – delta γ – gamma μL – microliter µM – micromolar % - percent sign A – Adenine
ATP – Adenosine Triphosphate Ala – Alanine
Arg – Arginine Asn – Asparagine Asp – Aspartic acid
BioFIG – Center for Biodiversity, Functional and Integrative Genomics BLAST – Basic Local Alignment Search Tool
bp – base pair C – Cytosine
c. – complementary DNA
CFTR - Cystic Fibrosis Conductance Regulator CHLN – Centro Hospitalar de Lisboa Norte CHLO – Centro Hospitalar de Lisboa Oriental cm2 – square centimetre
COXI – Cytochrome Oxidase I
Cx26 – Connexin-26 Cx30 – Connexin-30 Cys – Cysteine dB – decibel
DFNB1 – autosomal recessive nonsyndromic deafness DNA – Deoxyribonucleic Acid
EDTA - Ethylenediamine Tetraacetic Acid ENT – Ear, Nose and Throat
ER – Endoplasmic Reticulum G – Guanine
GJB2 – Gap junction Beta-2 GJB6 – Gap junction Beta-6
Gln – Glutamine Glu – Glutamic acid Gly – Glycine GTP – Guanosine triphosphate H2O – Water H2Od – Distilled water His – Histidine HL – Hearing Loss Hz – Hertz Ile – Isoleucine
IVS – Intervening Sequence K+ - potassium ion
kb – kilobase kDa – kilo Dalton Leu – Leucine
LFSNHL – Low Frequency Sensorineural Hearing Loss Lys – Lysine m. - mitochondrial M – Molar Met – Methionine mg – milligrams min – minute mL – milliliter mm2 – square millimeter mRNA – messenger RNA mtDNA – mitochondrial DNA NaCl – Sodium Chloride
NCBI – National Center for Biotechnology Information p – position on the chromosome’s short arm
p. – protein
PCR – Polymerase Chain Reaction Phe – Phenylalanine
Pro – Proline
PS – Pendred Syndrome
PSDM – PCR-Site Directed Mutagenesis q – position on the chromosome’s long arm rpm – revolutions per minute
s – second
SDS - Sodium Dodecyl Sulfate Ser – Serine
SLC26A4 - Solute Carrier Family 26, member 4
STAS - Sulfate Transporter and Antisigma factor antagonist T – Timine TBE – Tris/Borate/EDTA Thr – Threonine Tm – melting temperature Tris/HCl – Tris(hydroxymethyl)aminomethane-Hydrochloride Trp – Tryptophan
tRNA – transfer RNA Tyr – Tyrosine U – Unit Val – Valine
WFS1 – Wolfram Syndrome 1
Abstract
About 2/1000 new born present hearing loss (HL) and of these, approximately 50% can be genetically explained.
In this work, 70 Portuguese probands presenting nonsyndromic sensorineural HL were investigated for the presence of mutations in the DFNB1 locus. Mutation c.35delG, in the
GJB2 gene, was first screened by restriction analysis. Two individuals were homozygous
(2,9%) and five (7,1%) heterozygous for this mutation. The GJB2 coding exon was then analysed by sequencing in cases found to be negative or heterozygous for the c.35delG mutation. A novel mutation, p.Leu213X, was identified. This mutation wasn’t present in a random control sample of 480 individuals from the Portuguese population previously sequenced. Along with the two c.35delG homozygous, the genetic cause for HL due to GJB2 mutations could be determined for six more individuals, representing a total of 11,4% of the probands. The most frequent GJB6 deletions were further screened by multiplex PCR in individuals negative or monoallelic for GJB2 mutations. No mutation was found.
Three common mitochondrial DNA mutations associated to HL were investigated by enzymatic restriction in 143 families, previously screened for DFNB1 and compatible with maternal inheritance. Only the m.1555A>G mutation was found, in one of the analysed families.
Two cases of low-frequency sensorineural HL were investigated regarding the most relevant exons of WFS1 gene. A novel mutation, p.Asp171Asn, was found in exon 5 of WFS1 gene. This mutation wasn’t present in 100 Portuguese hearing controls later sequenced.
Two cases of Pendred Syndrome were studied. No mutation was found in the relevant exons analysed. So, the genetic cause couldn´t yet be determined.
This study represents a contribution for extending the knowledge on hereditary HL in the Portuguese population, either in the identification of the genetic etiology in affected families, or in the development of more accurate diagnostic protocols.
Resumo
Cerca de 2/1000 recém-nascidos apresentam surdez e, nestes, aproximadamente 50% têm uma causa genética.
No presente trabalho, foram estudados 70 probandos Portugueses com surdez não- sindrómica neurosensorial. A análise iniciou-se com o estudo da mutação c.35delG no gene
GJB2 através de restrição enzimática. Dois indivíduos apresentaram esta mutação em
homozigotia (2,9%) e cinco em heterozigotia (7,1%). O exão codificante de GJB2 foi analisado por sequenciação em todos os casos negativos ou heterozigóticos para c.35delG. Uma nova mutação, p.Leu213X, foi descoberta. Esta mutação não foi encontrada numa amostra aleatória de 480 indivíduos da população Portuguesa. Juntamente com os indivíduos homozigóticos para c.35delG, a causa genética da surdez devida a GJB2 foi determinada para mais seis indivíduos, representando no conjunto 11,4% dos probandos. As deleções mais frequentes no gene GJB6 foram investigadas por PCR multiplex em indivíduos monoalélicos ou negativos para GJB2. Nenhum dos indivíduos apresentou deleções em GJB6.
As três mutações mais comuns associadas a surdez no DNA mitocondrial foram investigadas por restrição enzimática em 143 famílias previamente testadas para GJB2 e compatíveis com hereditariedade materna. Apenas numa família foi encontrada a mutação m.1555A>G.
Dois casos de surdez neurosensorial com perdas nas baixas frequências foram estudados através da análise dos exões mais relevantes do gene WFS1. Uma nova mutação, p.Asp171Asn, foi encontrada no exão 5 deste gene. Esta mutação não foi encontrada em 100 controlos Portugueses ouvintes posteriormente sequenciados.
Dois casos esporádicos de Síndrome de Pendred foram estudados mas nenhuma mutação foi encontrada nos exões analisados, pelo que a causa genética não pôde ser ainda determinada.
Este estudo constituiu uma contribuição para aprofundar o conhecimento sobre a surdez hereditária na população Portuguesa. A importância do estudo genético fica aqui patente, quer na identificação da etiologia genética da surdez em famílias afectadas, quer no desenvolvimento de protocolos de diagnóstico mais eficazes.
Resumo alargado
A surdez representa a deficiência sensorial mais frequente na população humana, apresentando uma incidência de aproximadamente 2 em cada 1000 nascimentos. A surdez devida exclusivamente a causas genéticas tem tendência a aumentar nos países desenvolvidos, onde já cerca de 50% dos casos são devidos a surdez hereditária, em consequência da melhoria dos cuidados de saúde e diminuição da incidência dos factores ambientais. Dentro destes, em 70% a surdez é o único sintoma clínico (surdez não-sindrómica) e nos restantes 30% a surdez pode estar associada a outras desordens clínicas, sendo designada por surdez sindrómica. As formas de surdez não sindrómica autossómicas recessivas são geneticamente heterogéneas e constituem a forma mais comum da perda auditiva hereditária, contabilizando cerca de 75-85% dos casos. Este número é seguido pela hereditariedade autossómica dominante (12-13%) e pela hereditariedade mitocondrial ou ligada ao cromossoma X, que representa 2-3% dos casos de surdez genética.
Os mais recentes avanços na área da genética molecular permitiram associar a surdez a mais de 400 síndromes, ao mesmo tempo que mais de 140 loci relacionados com a surdez não-sindrómica foram mapeados. Até à data, cerca de 60 genes foram já associados com a perda auditiva.
O gene GJB2 (Gap Junction β-2) forma com o gene GJB6 (Gap Junction β-6) o locus DFNB1. Estes genes localizam-se no cromossoma 13q11-q12 e codificam para as proteínas conexina-26 (Cx26) e conexina-30 (Cx30), respectivamente. As proteínas conexinas possuem uma localização transmembranar e podem associar-se num hexâmero formando conexões (Martínez, 2009). Os conexões são designados homoméricos, quando são constituídos por um único tipo de conexina ou heteroméricos, quando são formados por diferentes tipos de conexinas. As gap junctions resultam da associação entre si de conexões na superfície de células adjacentes e são designadas de homotípicas, quando são formadas por conexões semelhantes ou heterotípicas, quando são formadas por conexões heteroméricos. As células ligadas por gap junctions usam este tipo de canais para transferir iões e outras pequenas moléculas entre si. Na cóclea, a Cx26 e a Cx30 encontram-se co-localizadas e co-expressas, contribuindo para a homeostasia coclear. A manutenção desta homeostasia é conseguida através da re-circulação de iões K+ na endolinfa após estimulação das células ciliadas do ouvido interno.
O gene GJB2 contém cerca de 5500 pb e um único exão codificante, num total de dois exões. Mais de 100 mutações foram já identificadas neste gene e, dependendo da população,
contribuem para cerca de 10-40% dos casos de surdez hereditária, representando a causa mais frequente de surdez hereditária não-sindrómica autossómica recessiva. A mutação mais comum na população Caucasóide é a c.35delG, que resulta numa terminação precoce da cadeia polipeptídica da Cx26, após uma alteração na grelha de leitura que conduz a um codão STOP prematuro. A frequência alélica para esta mutação foi estimada em 2,5% na população Caucasóide geral, num total de 3% de frequência alélica para todas as mutações em GJB2. A mutação c.167delT, por sua vez, possui uma elevada frequência na população de Judeus Ashkenazi (7,5%) e a mutação c.235delC é a mais frequente nas populações Coreanas e Japonesas, com frequências alélicas estimadas em 0,5 e 1%, respectivamente.
O gene GJB6, contrariamente ao gene GJB2, possui até hoje poucas mutações descritas, apenas seis, que apresentam um padrão de hereditariedade recessiva. Dessas seis mutações, quatro são deleções. As mais frequentes são as deleções Δ(GJB6-D13S1830) e Δ(GJB6-D13S1854), que, ao mesmo tempo que truncam o gene GJB6, inactivam o gene
CRYL1, eliminando a região entre estes. O gene CRYL1 codifica para a λ-cristalina e até à
data, nenhum caso de indivíduos que possuam quer a deleção Δ(GJB6-D13S1830), quer a deleção Δ(GJB6-D13S1854), foi reportado juntamente com desordens oculares.
A transmissão do DNA mitocondrial, através de herança materna, também contribui para casos de surdez não-sindrómica quando na presença de determinadas mutações. Uma destas mutações, m.1555A>G, constitui-se como uma causa comum de surdez não-sindrómica familiar. Esta mutação conduz a uma alteração na conformação do gene 12S
rRNA, tornando a sua estrutura semelhante ao gene ribossomal das bactérias. Assim, a
exposição a antibióticos aminoglicósidos, como a estreptomicina, torna-se ototóxica, podendo conduzir a um fenótipo de surdez acentuado.
As perdas auditivas nas altas frequências são relativamente comuns, ao contrário de perdas auditivas que afectem predominantemente as frequências abaixo dos 2000Hz. Devido à manutenção da audição nas altas frequências, os indivíduos que sofrem de perdas auditivas nas baixas frequências retêm a capacidade da percepção da fala, pese embora a presbiacúsia ou a exposição a ruídos elevados possam causar perdas auditivas nas altas frequências mais tarde nas suas vidas. A maioria dos casos reportados de perdas auditivas nas baixas frequências é devida a mutações dominantes no gene WFS1, que codifica para a wolframina, uma glicoproteína transmembranar localizada no retículo endoplasmático.
Entre os diferentes síndromes em que a surdez se encontra associada a outras anomalias, destaca-se o Síndrome de Pendred. Este síndrome, autossómico recessivo,
representa a forma mais comum de surdez sindrómica e apresenta uma incidência de cerca de 7,5 a 10 casos por cada 100000 indivíduos. Os indivíduos afectados com Síndrome de Pendred possuem mutações bialélicas no gene SLC26A4 e apresentam, para além de surdez neurosensorial, defeitos na tiróide e malformações no ouvido interno, que podem originar displasia de Mondini.
O principal objectivo deste trabalho foi determinar a causa genética da surdez apresentada por pacientes Portugueses com surdez neurosensorial não-sindrómica. Casos esporádicos de Síndrome de Pendred e de perdas auditivas nas baixas frequências foram também estudados.
Setenta probandos com fenótipo de surdez neurosensorial não-sindrómica foram primeiramente testados para a presença da mutação c.35delG, por restrição enzimática. Esta mutação foi encontrada em homozigotia em dois (2,9%) probandos e em heterozigotia em cinco (7,1%) probandos. Para os 2 indivíduos homozigóticos para a mutação c.35delG estava nesta altura encontrada a causa genética do seu fenótipo de surdez. Para os restantes 68 indivíduos, negativos (n=63) e heterozigóticos (n=5) para a mutação c.35delG, foi feita a sequenciação para o exão codificante do gene GJB2. Esta sequenciação revelou a presença de 14 diferentes variantes nos indivíduos considerados. Para seis casos (8,8% 6/68) a causa genética da surdez foi determinada através de sequenciação automática, enquanto para os restantes permaneceu inconclusiva. A causa genética da surdez pôde ser estabelecida para: um probando heterozigótico composto com as mutações p.Met34Thr e p.Iso140Ser; um probando com heterozigotia composta envolvendo as mutações c.333-334delAA e p.Leu213X; um probando homozigótico para a mutação p.Trp24X; um probando com heterozigotia composta envolvendo as mutações p.Met34Thr e p.Arg184Pro; um probando heterozigótico composto c.35delG e p.Trp172X e, um probando apresentando uma heterozigotia composta com as mutações p.Val37Ile e p.Asn206Ser. No conjunto, 11,4% dos probandos (8/70) apresentam surdez associada ao gene GJB2. Durante esta etapa, uma nova mutação nunca antes descrita foi reportada, p.Leu213X, com localização no C-terminal da Cx26. Esta mutação não foi encontrada numa amostra aleatória de 480 indivíduos Portugueses que tinha sido previamente sequenciada para o exão codificante do gene GJB2.
Para os probandos monoalélicos ou sem mutações em GJB2 procedeu-se à análise por PCR multiplex das deleções mais frequentes no gene GJB6. Nenhuma das deleções testadas foi encontrada.
Relativamente à análise das três mutações mais comuns no DNA mitocondrial associadas a surdez, dos 143 indivíduos analisados apenas um probando foi diagnosticado com heteroplasmia para a mutação m.1555A>G. Posteriormente, a análise por restrição enzimática à irmã e à mãe revelou neles também a presença desta mutação.
No âmbito deste trabalho, foram também analisadas duas famílias que apresentavam fenótipo de perdas auditivas nas baixas frequências. Foram sequenciados os exões 4, 5, 6 e 8 do gene WFS1 onde se localizam, até à data, a maioria das mutações patogénicas descritas. Para uma das famílias este rastreio foi inconclusivo. No que diz respeito ao segundo caso, foi encontrada uma nova mutação, p.Asp171Asn, não descrita até à data. Esta mutação, presente em heterozigotia num indivíduo não foi encontrada numa amostra de 100 indivíduos ouvintes Portugueses que foram sequenciados para o exão 5 do gene WFS1.
Dois probandos pertencentes a duas diferentes famílias Portuguesas foram encaminhados para este estudo com a indicação de um fenótipo compatível com Síndrome de Pendred. Para estes indivíduos foi feito o estudo por sequenciação automática dos exões 6 e 10 e das regiões IVS8 e IVS14 do gene SLC26A4. Não tendo sido encontrada nenhuma variante, a causa genética do fenótipo apresentado por estes indivíduos não ficou elucidada. No entanto, no estudo da região IVS14, onde parte do exão 15 foi também analisada, houve a suspeita de uma nova mutação, p.Ser552Gly, ter sido encontrada. Este facto causou estranheza, pois esta nova mutação surgira nos dois indivíduos de diferentes famílias em homozigotia. Um novo protocolo de diagnóstico molecular foi desenhado e clarificou-se que se tratou de um artefacto induzido ou na reacção de PCR ou na sequenciação automática, não constituindo uma nova mutação.
O estudo molecular reveste-se de grande importância, quando se considera a probabilidade de uma futura gravidez vir a gerar novamente um descendente com perturbações auditivas. Uma análise cuidada do heredograma familiar, juntamente com a análise genética, constitui-se também de grande utilidade, na medida em que outros membros da família podem vir a beneficiar de um teste semelhante, se a causa genética da surdez for familiar. Um diagnóstico genético atempado pode elucidar a causa e a evolução da surdez (progressividade ou não). Com base nesse resultado, melhores estratégias terapêuticas podem ser equacionadas, contribuindo decisivamente para uma melhor saúde auditiva e uma melhor qualidade de vida dos indivíduos afectados e, também, com menores custos associados.
1. Introduction
1.1. Genetic Hearing Loss
Hearing loss (HL) is the most frequent disability of the human senses (Pollak et al, 2007). Clinically relevant sensorineural HL (SNHL) is present in at least 2 per 1000 new born at birth, rising to at least 2,7 per 1000 infants by the age of four (Petersen et al, 2012). About 50-70% of hearing impaired children have a monogenic cause for their deafness (Matsunaga, 2009).
Hereditary HL can be subdivided into two types: syndromic or nonsyndromic. The syndromic type is associated with other distinctive clinical features beyond deafness and accounts for 30% of hereditary congenital HL. The nonsyndromic type, in which HL is the only clinical manifestation, represents the other 70% (Matsunaga, 2009).
Regarding nonsyndromic HL, autosomal recessive is the most frequent inheritance pattern, accounting for 75-85% of the cases (Snoeckx et al, 2005; Ibrahim et al, 2011). It is followed by dominant pattern (12-13%) and X-linked or mitochondrial inheritance that accounts for 2-3% of the cases (Ibrahim et al, 2011).
More than 140 genetic loci that have been associated to nonsyndromic HL were mapped, being identified to date about 60 genes (Minami et al, 2012).
1.2. DFNB1 locus
At chromosome 13q11-q12, DFNB1 locus comprises Gap Junction β-2 (GJB2) and
Gap Junction β-6 (GJB6) genes that encode for connexin 26 (Cx26) and connexin 30 (Cx30),
respectively (Rodríguez-Paris et al, 2011).
Connexin are transmembrane proteins, being constituted by intracellular amino- and carboxy-termini and four transmembrane domains. Connexons are transmembrane hexameric gap junction hemi-channels composed by six connexin proteins (Martínez et al, 2009). Connexons can be classified into homomeric, when they are made up of a single type of connexin or heteromeric, when they are formed by different connexin proteins (Erbe et al, 2004; Tang et al, 2006). Connexons embedded in the surfaces of adjacent cells can associate and form intercellular channels. Intercellular channels then cluster and form gap junctions. Gap junction channels can be homotypic, when are composed of similiar connexons or heterotypic, when made up of different connexons (Martínez, 2009). Cells connected by gap junctions use the channels to transfer ions and other small molecules across cell membranes (Tang et al, 2006).
Cx26 and Cx30 are co-localised and co-expressed in the cochlea, an organ in the inner ear, where they contribute to cochlear homeostasis, since they are thought to provide the recirculation of K+ ions to the endolymph after hair cell’s sound stimulation (fig.1) (Rodríguez-Paris et al, 2011; Erbe et al, 2004).
1.2.1. GJB2 gene
GJB2 gene (fig.2) has 5500 bp and a single coding exon, in a total of two exons (Falah et al, 2011). More than 200 different pathogenic mutations were identified in this gene that
account for 10-40% of congenital HL depending of ethnicity, being the most frequent cause of nonsyndromic autosomal recessive hereditary HL (Tang et al, 2006; Falah et al, 2011).
The majority of mutated Cx26 alleles among Caucasoids worldwide are due to a deletion of a guanine within a string of six guanines at nucleotide 35 (c.35delG) that results in a premature chain termination (Gasparini et al, 2000). Carrier frequency for this allele was estimated to be 2,86% in countries from southern Europe and 1,27% in countries from central and northern Europe (Gasparini et al, 2000). In United States Caucasoids c.35delG carrier rate was found to be 2,5% (Green et al, 1999). In Portugal, the carrier rate for c.35delG in general population was estimated to be 0,88% (Chora et al, 2011). The absence of the c.35delG mutation in other populations, like North American Blacks, Egyptians and Yemenite Jews, for example, could be explained as a consequence of a single origin, somewhere in Europe or in the Middle East (Gasparini et al, 2000). The high frequencies observed for c.35delG carriers
Figure 1 – Schematic representation of the cochlea, showing the location of its different structures. The K+ recycling pathway is indicated. Adapted from Jentsch, 2000.
in some Caucasoid population suggest either a founder effect or a selective advantage for heterozygotes, or even a combination of both (Gasparini et al, 2000). It was hypothesized (Gasparini et al, 2000) that since Cx26 is expressed among a variety of tissues, it is conceivable that a putative heterozygote advantage is related to a function of Cx26 in one of these tissues, but clearly not the cochlea. This advantage could be associated to specific functions of gap junctions and could be involved in climate, food, toxic factors and infectious agents, among others, reflecting geographic and cultural different conditions that could influence the frequency of c.35delG allele (Gasparini et al, 2000).
GJB2 mutations associated with other specific ethnic groups include c.167delT, with a
carrier rate of 7,5% in Ashkenazi Jews and c.235delC, with a carrier rate of 0,5% to 1% in Korean and Japanese populations (Erbe et al, 2004).
1.2.2. GJB6 gene
Whilst many mutations have been described in GJB2 gene, so far only six mutations with a recessive pattern of inheritance were reported in GJB6 gene or the region upstream causing deafness (Ballana et al, 2012). Four of them are deletions, as shown in figure 2. The most frequent deletions are Δ(GJB6-D13S1830) and Δ(GJB6-D13S1854), which truncate the
GJB6 gene. Both Δ(GJB6-D13S1830) and Δ(GJB6-D13S1854) mutations inactivate the CRYL1 gene and eliminate the sequence between GJB6 and CRYL1, where no additional
genes have been reported so far. λ-crystallin is the product of CRYL1 gene and no contribution of λ-crystallin to DFNB1 HL was found to date. Additionally, no individual carrying either Δ(GJB6-D13S1830) or Δ(GJB6-D13S1854) was found to present any eye disorder (del Castillo et al, 2002; del Castillo et al, 2005). The other two deletions affecting
GJB6 gene are private mutations, one of which (>920 kb) deletes both GJB2 and GJB6 genes
and the other one [(Δ(chr13:19,837,343–19,968,698))] does not affect either gene and is located upstream of GJB6 (Rodríguez-Paris et al, 2011).
Figure 2 – Map of the region on chromosome 13q11-12 showing GJB2, GJB6 and CRYL1 genes, relatively to centromere. The regions disrupted by the deletions are indicated. Adapted from Rodríguez-Paris et al, 2011.
1.3. Low Frequency Sensorineural Hearing Loss
Low frequency SNHL (LFSNHL) is an unusual form of HL, in which frequencies at 2000 Hz and below are predominantly affected. Due to maintenance of high frequency hearing, LFSNHL patients retain understanding of speech, although presbycusis or noise exposure may cause high frequency loss later in their life (Bespalova et al, 2001). Four loci, DFNA1, DFNA6, DFNA14 and DFNA38, are reported as being associated with LFSNHL (Bespalova et al, 2001; Young et al, 2001).
Most of the families with LFSNHL carry mutations in WFS1 gene that maps to chromosome 4p16 and has a coding transcript of 2673 bp. WFS1 gene has 8 exons, of which only the last seven are coding (Gürtler et al, 2004; Minami et al, 2012). The product of WFS1 is wolframin, a membrane glycoprotein that is located primarily in the endoplasmic reticulum (ER). Its expression in the human cochlea remains unknown (Gürtler et al, 2004). However, its location in the ER suggests a possible role for wolframin in ion homeostasis retained by the canalicular reticulum, a specialized form of ER (Minami et al, 2012). Studies of functional analysis suggest that the autosomal dominant pattern of LFSNHL is due to reduced amount of wolframin (Gürtler et al, 2004).
Mutations in WFS1 gene can cause autosomal dominant LFSNHL (DFNA6/14/38) (Bespalova et al, 2001; Cryns et al, 2003). They are also associated with Wolfram Syndrome (WS), an autosomal recessive syndrome, characterized by insulin-dependent diabetes mellitus and bilateral progressive optic atrophy, usually presenting in childhood or early adult life (Cryns et al, 2003; Aloi et al, 2012).
1.4. Syndromic Hearing Loss
The syndromic type of hereditary HL includes about 400 syndromes, such as Pendred Syndrome (PS), the most frequent form of syndromic HL. It accounts for 4–10% of the inherited cases (Fraser, 1965; Illum et al, 1972; Reardon et al, 1997), and is inherited in an autosomal recessive pattern, with an incidence estimated to be as high as 7,5 to 10 in 100000 individuals (Reardon et al, 1997; Fraser, 1965).
PS is characterized by SNHL, goiter and a partial defect in iodide organification. These features are generally accompanied by malformations of the inner ear, ranging from enlarged vestibular aqueduct (EVA) to Mondini dysplasia (Pera et al, 2008).
The clinical features observed in PS are consequence of biallelic mutations in the
SLC26A4 gene. This gene, containing
21 exons, is located on chromosome 7 (7q22.3-q31.1) and codes for the multifunctional anion exchanger pendrin (fig.3) (Pera et al, 2008; Bizhanova, 2010). Pendrin is a 73 kDa membrane protein that belongs to the solute carrier family 26A. It is
comprised of 780 amino acids and is predicted to have 12 putative transmembrane domains, being both the amino- and carboxi-termini located on the cytosol (Royaux et al, 2000; Gillam
et al, 2004). Pendrin has a sulfate transporter domain and a sulfate transporter and antisigma
factor antagonist (STAS) domain. The last one has been suggested to play a role in nucleotide binding or to interact with other proteins, such as cystic fibrosis conductance regulator (CFTR) (Bizhanova et al, 2010).
1.5. Mitochondrial Hearing Loss
Mitochondria are organelles involved in oxidative metabolism, ion homeostasis, signal transduction and apoptosis. These cellular organelles contribute to pathogenicity by a variety of mechanisms involving maternally inherited diseases due to mutations in mitochondrial (mt) DNA, as well as Mendelian-inherited diseases resulting from mutations in nuclear genes required for mitochondrial function (Raimundo et al, 2012). The mtDNA is a double-stranded circular genome composed of 16569 bp that codes for 13 subunits of respiratory complexes (Angulo et al, 2011). Normal cochlear function requires a very high rate of ATP production and mtDNA mutations have often been found to cause hearing deficiencies either in syndromic and nonsyndromic forms of HL (Guan et al, 2008). Up to now, eight mutations in mtDNA have been clearly associated to HL phenotype (Mitomap, 2012). One of these, the m.1555A>G mutation in 12S rRNA gene (fig.4) is a common cause of familiar nonsyndromic post-lingual SNHL. The HL phenotype due to this mutation is significantly exacerbated after exposure to aminoglycosides, in particular streptomycin, because this mutation alters the 12S
rRNA gene conformation, making it similar to the bacterial ribosomal gene, thus enhancing
aminoglycoside binding and its toxic effects in the ear (Angulo et al, 2011).
Figure 3 – Schematic representation of pendrin protein. Figure adapted from Bizhanova, 2010.
The m.7445A>G mutation in the
tRNASer(UCN) gene, also called COXI (Cytochrome
oxidase I) gene, is another mtDNA mutation
associated with HL (Tekin et al, 2003). It occurs in the immediately adjacent nucleotide to the 3’-end of tRNASer(UCN) and previous studies indicate that it influences the normal processing of the light strand polycistronic mRNA, being the primary defect a significant decrease in serine tRNA level and protein synthesis rate in mitochondria (Guan et al, 2008). The m.7511T>C mutation also occurs in
tRNASer(UCN) gene and was associated to nonsyndromic deafness in several families from different ethnic groups (Zheng et al, 2012). This mutation is responsible for the substitution of a highly conserved A-U to a G-U base pairing on the 5’ side of the acceptor stem of the tRNASer(UCN) (Zheng et al, 2012). Mutations in the tRNASer(UCN) gene often occur at high degrees of heteroplasmy or in homoplasmy, indicating a high threshold for pathogenicity (Zheng et al, 2012). In figure 5 it is shown a schematic representation of the human mtDNA.
A B
Figure 4 – The coding region of the 12S
rRNA human mtDNA gene. A – wild-type
and B – m.1555A>G mutation. Adapted from Böttger, 2010.
Figure 5 – Schematic representation of human mtDNA showing the loci involving in mtDNA associated diseases. The genes 12S and COX I are emphasized with red boxes. Adapted from Griffiths, 2004.
2. Objectives
The main objective of this work was to investigate the genetic etiology of the hearing- -impaired phenotype presented by Portuguese patients with SNHL.
The specific objectives were:
To analyse the coding exon of the GJB2 gene;
To perform the screening of the most common deletions in GJB6 gene;
To analyse the most frequent mutations in mtDNA associated with deafness;
To screen the WFS1 relevant exons in LFSNHL cases;
3. Materials and Methods
3.1. Sample characterization
A total of 152 Portuguese hearing impaired individuals were selected for this study: 70 patients referred by the ENT Department, CHLO - Hospital Egas Moniz; 81 cochlear implanted individuals referred by the ENT Department, CHC, EPE - Hospital dos Covões, Coimbra, previously analysed for the GJB2 coding region and identified with negative results (Chora et al, 2010) and one individual from ENT Department, CHLN - Hospital Santa Maria. The majority of these patients presented bilateral nonsyndromic SNHL. Two individuals presenting LFSNHL and two cases of syndromic HL were also selected and included in this study. Blood samples were always codified with a blind code, composed of letters (family code) plus numbers (individual code), prior to the DNA extraction and sample manipulation. The personal information of each individual was carefully stored with restricted access. A detailed clinical history of each proband was taken to ensure that the HL was not a result of infection, acoustic trauma, ototoxic drugs or premature birth. For some patients, however, family histories weren’t possible to establish. Written informed consent was obtained from all the participants in this study.
3.2. Audiologic examination
The probands and some members of their families underwent otoscopic and audiometric examinations by using age-appropriate methods. Pure tone audiometry was obtained by the clinician’s teams of the considered Hospitals listed in 3.1 section in a sound proof room according to current clinical standards. The level of HL was classified following the European Working Group on Genetics of Hearing Impairment as slight (21–40 dB), moderate (41–70 dB), severe (71–95 dB), or profound (>95dB), from an average at 500, 1000, 2000 and 4000 Hz in the better ear.
3.3. Extraction of genomic DNA from blood samples
The protocols for the extraction of genomic DNA from blood samples are presented in section 1 of the supplementary data.
3.4. Amplification by Polymerase Chain Reaction (PCR)
PCR was performed in order to amplify specific studied DNA regions described below. In each PCR reaction a negative control, in which DNA wasn’t added, was used to confirm the absence of contaminants in the reagents. All primers used are listed in table 5 of
the supplementary data. PCR reactions were made in Biometra T1 Thermocycler and PCR products were always run in a 1% agarose gel in TBE 0,5x by electrophoresis, as described in section 2 of the supplementary data.
3.5. Screening of c.35delG mutation
PCR-mediated Site Directed Mutagenesis was the method chosen to test for the c.35delG mutation in the GJB2 gene. It consisted in originating another restriction site for the restriction enzyme BslI (New England Biolabs) in the presence of the mutation, after amplification with a modified specific primer. The PCR reaction was performed for the test fragment (called δ) and for an internal control of the restriction (called γ). PCR reaction for δ fragment used 22BF 80µM and 22BR 20µM primer pair and is described in detail in table 1 of the supplementary data. The γ fragment was amplified using FP 10µM and RP 10µM primer pair following the conditions listed in detail in table 1 of the supplementary data. PCR programme for the amplification of δ and γ fragments is shown in table 2 of the supplementary data.
3.5.1. Restriction with BslI
After PCR amplification, the screening of c.35delG mutation was made with the restriction enzyme BslI as referred. The γ fragment, which also has a recognition site for BslI, was used as an internal control of the restriction. To maintain DNA concentration in the reaction mix, 8µL of δ product + 8µL of γ product were concentrated during 7 min at 45ºC on DNA SpeedVac. The total volume of each sample restriction and the incubation conditions are indicated in table 4 of the supplementary data.
The δ fragment is not digested in the absence of c.35delG mutation but, in the presence of the mutation, this fragment of 207 bp is digested into 181 + 26 bp. The control γ fragment has 2 overlapping cutting sites. So, when the enzyme cuts one of them, the other disappears, and the 153 bp control fragment can be digested into 97 + 56 bp or 99 + 54 bp.
3.6. Amplification of GJB2 coding exon
PCR amplification of the coding exon of the GJB2 gene is listed in detail in table 1 of the supplementary data. Standard PCR programme was used following the conditions shown in table 2 of the supplementary data. The amplified PCR product had 928 bp. GJB2 PCR products from the probands heterozygous or negative for c.35delG mutation and, when necessary from parents and siblings, were automatically sequenced after purification with
JETQUICK PCR Product Purification Spin Kit (Genomed) according to manufacturer’s instructions.
3.7. Screening for GJB6 deletions
The GJB6 deletions del(GJB6-D13S1830) and del(GJB6-D13S1854) can be detected by multiplex PCR (del Castillo, 2005). Three different primer pairs were used, which amplified according to their proximity, depending on the presence of the deletions. The Cx30Ex1A and Cx30Ex1B primer pair amplifies a 333 bp fragment in the presence of a wild- -type allele. The Del BK1 and Del BK2 primer pair amplifies a 564 bp fragment in the presence of a mutated allele with del(GJB6-D13S1854) deletion. The GJB6-1R and BKR-1 primer pair amplifies a 460 bp fragment in presence of del(GJB6-D13S1830). PCR amplification of the GJB6 screened deletions followed the conditions listed in detail in table 1 of the supplementary data. All primers were used at 10µM. The PCR program included a touchdown step, as described in table 3 of the supplementary data.
3.8. Mitochondrial DNA analysis
Screening of three known mutations: m.1555A>G, m.7445A>G and m.7511T>C was performed for all those cases compatible with maternal inheritance and which cause of deafness remained to be elucidated.
3.8.1. Detection of m.1555A>G mutation
This mutation can be detected by PCR with specific primers that amplify the 12S rRNA gene, followed by digestion with the restriction enzyme HaeIII (Promega). PCR mix is described in detail in table 1 of the supplementary data. The standard PCR program is listed in table 2 of the supplementary data. The PCR product had 339 bp.
The restriction enzyme HaeIII always recognizes a cleavage site in the mentioned amplified PCR area and cuts originating two fragments of 218 bp + 121 bp. The m.1555A>G mutation creates one other cleavage site and thus the enzyme cuts twice, originating three fragments of 218 bp + 91 bp + 30 bp. The restriction mix and the incubation conditions are indicated in table 4 of the supplementary data.
3.8.2. Detection of m.7445A>G and m.7511T>C mutations
The m.7445A>G and m.7511T>C mutations can be detected by PCR with specific primers that amplify part of the the tRNASer(UCN) gene, followed by digestion with the
is described in detail in table 1 of the supplementary data. The standard PCR program is also listed in table 2 of the supplementary data. The PCR product had 215 bp.
The m.7445A>G mutation is recognized by the loss of a cleavage site for the XbaI restriction enzyme, as compared to the wild-type. So, in the wild-type, XbaI always has a recognition site and cleaves, originating two fragments of 166 bp + 49 bp. The restriction mix and the incubation conditions are shown in table 4 of the supplementary data.
MboII restriction enzyme recognizes a cleavage site only in the wild-type. This
enzyme cuts in the absence of the mutation m.7511T>C, originating two fragments of 175 bp + 40 bp. The restriction mix and the incubation conditions are indicated in table 4 of the supplementary data.
3.9. Screening of WFS1 exons 4, 5, 6 and 8
For those individuals presenting LFSNHL, automatic sequencing of exons 4, 5, 6 and 8 of WFS1 gene was performed. PCR amplification mix is listed in table 1 of the supplementary data. The standard PCR programme is also shown in table 2 of the supplementary data. Exon 8 was amplified in two PCR reactions (designed in this study by 8a and 8b) because its large size is not compatible with good quality of automatic sequencing.
WFS1 PCR products of exons 4 (222 bp), 5 (225 bp), 8a (872 bp) and 8b (1096 bp) were
automatically sequenced after purification with JETQUICK PCR Product Purification Spin Kit (Genomed) according to manufacturer’s instructions. PCR products of exon 6 (186 bp) were automatically sequenced after purification with Zymoclean™ Gel DNA Recovery Kit (Zymoresearch) following manufacturer’s instructions.
One hundred normal hearing Portuguese controls were sequenced for the exon 5 of
WFS1 gene after purification with JETQUICK PCR Product Purification Spin Kit (Genomed).
3.10. Screening of SLC26A4 exon 6, exon 10, IVS8 and IVS14 regions
For those individuals presenting medical diagnostic of PS, automatic sequencing of exons 6 and exon 10, as well as IVS8 and IVS14 regions of SLC26A4 gene, was performed. PCR amplification of each region of SLC26A4 gene is shown in detail in table 1 of the supplementary data. In table 2 of the supplementary data is listed the standard PCR programme used. SLC26A4 PCR products of exons 6 (459 bp), exon 10 (606 bp), IVS8 (499 bp) region and IVS14 (185 bp) region were automatically sequenced after purification with JETQUICK PCR Product Purification Spin Kit (Genomed) according to manufacturer’s instructions.
Results and Discussion
4.1. GJB2 gene analysis
GJB2 analysis was the first step performed to determine the genetic cause of HL in
probands from 70 families recently referred to the Deafness Group of the BioFIG. This analysis started by screening c.35delG mutation in probands, since it represents the most frequent mutation in Cx26 gene in Caucasoid population (Zelante et al, 1997). Its detection was done for all the probands prior to the analysis of the GJB2 coding exon, performed for the patients shown to present only one or no c.35delG mutation.
4.2. Screening of the c.35delG mutation After restriction analysis for c.35delG mutation, seven of the 70 probands (10%) were found to be either heterozygous (7,1%) or homozygous (2,9%) for this mutation. An image of a c.35delG restriction analysis agarose gel is shown in figure 6.
One of the heterozygous proband was later shown, after sequencing of GJB2 coding exon (section 4.3.11), to have the
c.35delG mutation in compound
heterozygosity with another GJB2 mutation, p.Trp172X.
So, in the present study c.35delG
mutation explains the HL in three of the 70 patients (4,3%): two homozygous and one in compound heterozygosity with p.Trp172X mutation. The mutation c.35delG has been reported as the most common mutated allele found among Mediterranean HL families (Zelante et al, 1997). The results of this study agree with this finding, since c.35delG was the most common allele found among the studied probands, representing 9 alleles in a total of 140 (6,4%).
Figure 6 – Agarose gel from a c.35delG restriction analysis. Lanes: 1 – 1kb DNA plus ladder (Invitrogen); 2 – non-digested control; 3-4 and 7-8 – wild-type samples for c.35delG; 5 – heterozygous sample for c.35delG; 6 – homozygous sample for c.35delG.
4.3. Automatic sequencing of the GJB2 coding exon
Automatic sequencing of coding exon of GJB2 gene was performed for the probands from the remaining 68 families, heterozygous and negative for c.35delG mutation. When a mutation/variant was found, the parents and siblings DNA, if available, was also sequenced.
Fifty-six of the 68 sequenced probands (82,3%) were shown to be wild-type for GJB2 alleles. In 12 probands (17,6%, 12/68), Cx26 variants were found and in 6 families (8,8%, 6/68) the diagnosis for HL could be establish. The relevant results are described below, for each family concerned.
4.3.1. Case KQ
A single individual presenting SNHL was available for study. His GJB2 genotype was found to be [=] + [p.Lys224Gln]. The p.Lys224Gln (fig.1, supplementary data) mutation occurs at the intracellular C-terminal domain, which is involved in pH gating of Cx26 channel (Kelley et al, 1998). Due to the fact that this is a recessive mutation, no conclusion could be drawn as regards the etiology of the HL in this individual.
4.3.2. Family MV
This family is composed of two siblings presenting moderate SNHL and the normal hearing parents. The GJB2 genotype was found to be [p.Met34Thr] + [p.Val95Met] for one of the siblings (fig.7, II:2) and [=] + [p.Val95Met] for the other one (fig.7, II:1). The p.Val95Met variant (fig.2, supplementary data), occurring in intracellular loop (Martinez et al, 2009), was inherited from the mother who presents the
genotype [=] + [p.Val95Met]. The father was found to be a p.Met34Thr (fig.3, supplementary data) carrier.
The p.Val95Met is a controversial variant. Studies by Wang et al (2003) and Zhang et
al (2005), shows that the ionic permeability of p.Val95Met-Cx26 channels is not affected.
However, the same study by Zhang, and Beltramello et al (2005) conclude that the permeability to large molecules is impaired. The p.Val95Met was found in compound heterozygosity with other Cx26 mutations (c.35delG, p.Met34Thr and p.Leu90Pro) in several HL patients (Cryns et al, 2004; Snoeckx et al, 2005).
[=] + [p.Val95Met]
[=] + [p.Val95Met] [p.Met34Thr] + [p.Val95Met] [=] + [p.Met34Thr]
I:1 I:2
II:1 II:2
The p.Met34Thr variant, occuring in transmembranar domain 1 of Cx26, is a controversial variant, described as a recessive mutation by some authors (Snoeckx et al, 2005) and as a benign polymorphism by other authors (Feldmann et al, 2004). Data in literature are quite contradictory for p.Met34Thr variant, with some studies concluding that p.Met34Thr does not interfere with efficient formation of stable connexons (Oshima, 2003) and other studies concluding that the defect observed in the function of mutant protein is due to a shift in the gating response of the channel, resulting in a low conductance and permeability (Skerrett et al, 2004).
Due to the fact of both mutations being controversial and the affected son having the same genotype as his hearing mother, we cannot assume that the HL in this family is associated to the GJB2 genotypes observed.
4.3.3. Family NI
This family is composed of three normal hearing siblings (fig.8 II:1, II:2 and II:3) and their parents (fig.8 I:1 and I:2), having other relatives with deafness. The siblings and their mother presented the [=] + [p.Arg127His] GJB2 genotype. The variation p.Arg127His (fig.4, supplementary data) is localised to the intracellular loop 2 of Cx26 (Martinez et al, 2009).
The p.Arg127His variant was found to greatly impair the ability of Cx26 protein to form functional gap junctions, by reducing the channel permeability (Wang et al, 2003).
However, other study reported homozygous individuals for p.Arg127His that harboured normal hearing condition (Rouxet al, 2004). Moreover, additional data show that frequency
of p.Arg127His between normal hearing individuals and HL patients is not statistically different, demonstrate that p.Arg127His is effectively a polymorphism (RamShankar et al, 2003).
Mutation p.Arg127His is still controversial, but it is possible that it exerts a pathogenic effect depending on the environment and genetic background (Matos et al, 2010). In this family all the affected individuals carry the p.Arg127His. They may be coincidental carriers, but it is also possible that a noncoding GJB2/GJB6 mutation in trans with p.Arg127His might
Figure 8 – Heredogram of Family NI.
[=] + [=]
I:1 I:2
II:1 II:2 II:3
[=] + [p.Arg127His]
4.3.4. Case NP
A single individual presenting severe bilateral SNHL was available for study. His
GJB2 genotype was found to be [=] + [p.Gly160Ser]. The p.Gly160Ser mutation (fig.5,
supplementary data) localizes to the extracellular loop 2 of Cx26 and is described as a polymorphism (Tang et al, 2006). These data are not sufficient for the establishment of the HL genetic cause in this individual.
4.3.5. Case NT
A single individual presenting familiar cases of deafness was available for study. The +785 A>T and +792 C>T variations were detected on GJB2 gene in heterozygosity. Both variations were first reported by Tang and his collaborators (Tang et al, 2006), and were found at high frequence among the african-american population. This results aren’t strange considering the similarity of the genetic background of Portuguese population in relation with the African population. No genetic cause was yet determined for the HL in this individual.
4.3.6. Case NY
One individual presenting bilateral severe nonsyndromic SNHL was available for study. His GJB2 genotype was found to be compound heterozygous for the variants [p.Met34Thr]+[p.Ile140Ser] (Gonçalves et al, 2012, A). This individual has a sister with HL, of unknown genotype since she was not available for study. Mutation p.Ile140Ser (fig.6, supplementary data), localised to transmembrane domain 3 of Cx26, was first reported in 2005 (Snoeckx et al, 2005) as being recessive, and is not enough to explain the HL of this individual (Martinez et al, 2009; Snoeckx et al, 2005). Taking into account the role of p.Met34Thr variant described in 4.3.2 section, the conjugation of p.Ile140Ser with p.Met34Thr variant may justify the HL phenotype observed.
4.3.7. Family OB
This family is
composed of two hearing- impaired siblings (fig.9) with
moderate and severe
nonsyndromic SNHL (fig.10), respectively aged 11 (fig.9 II:2, fig.10D) and 15 (fig.9
[c.333-334delAA] + [p.Leu213X] [c.333-334delAA] + [p.Leu213X]
I:1 I:2
II:1 II:2
Figure 9 – Heredogram of Family OB.
II:1, fig.10C) (proband) years and their normal hearing parents (fig.9 I:1 and I:2 and fig.10 A and B, respectively). Both siblings shared the same GJB2 genotype [c.333-334delAA] + [p.Leu213X], being the deletion c.333-334delAA inherited from the mother. The p.Leu213X is a novel variant, identified for the first time in this study and is present in both siblings (Gonçalves et al, 2012, accepted paper, B). This mutation was later found in heterozygosity in their father.
The recessive deletion c.333-334delAA (fig.11,B) (Kelley et al, 1998), localised to the intracellular loop of Cx26, causes a frameshift which results in chain termination after an additional novel amino acid, truncating about half of the protein. It is the first time that this mutation is reported in Portuguese HL patients.
The novel recessive mutation p.Leu213X (fig.11,D) creates a premature STOP codon by changing the codon 213 (TTG) which codes for a leucine, to a STOP codon (TAG). This mutation leads to the deletion of the last 14 amino acids of the protein. The Leu213 amino acid residue is localised to the C-terminus domain of the Cx26 protein. The residues in the intracellular loop region and C-terminus are very different among different connexins and are hence thought to be responsible for regulation, thus imparting unique properties to the various connexin molecules (Mani et al, 2009). The p.Leu213X mutation wasn’t present in 480
Figure 10 – Representative audiograms of Family OB. Individual I:1 (A), individual I:2 (B), individual II:1 (C) and individual II:2 (D) showing pure-tone audiometry results for air conduction bilaterally. Circles in blue represent the right ear; crosses in red represent the left ear.
A B
Portuguese individuals from a random control sample that were previously sequenced for
GJB2 gene (Chora et al, 2011). Future functional studies are necessary to characterize this
novel mutation p.Leu213X.
It can be conclude that the etiology of deafness in these siblings is most likely due to the GJB2 genotype involving the c.333-334delAA deletion and the novel p.Leu213X mutation in compound heterozygosity.
4.3.8. Family OF
This gypsy family is composed of five individuals, three siblings and two related individuals. A previously described variation in the nucleotide position -40 (from the ATG initiation codon) was found in all the individuals of this family. This variation, c.-22-18T>A, was found in homozygosity in the proband and his affected siblings. It was also found in heterozygosity in the two relatives. This variation was reported previously (Tang et al, 2006), but in heterozygosity. So, to our knowledge, this is the first time that this variation is reported in homozygosity.
One of the two relatives carrying the c.-22-18T>A variation, was also found to be a carrier for the GJB2 mutation p.Trp24X. The role of this mutation will be described next in 4.3.9 section.
The second relative was found to present, besides the c.-22-18T>A heterozygosity, the p.Phe83Leu polymorphism. The polymorphism p.Phe83Leu localised to the transmembranar
Figure 11 – Electrophoretograms of c.333-334delAA and p.Leu213X mutations, respectively. A and B – wild-type and c.333-334delAA mutation in heterozygosity, respectively. C and D – wild-type and p.Leu213X mutation in heterozygosity, respectively.
domain 2 of Cx26 and was previously reported as a non-pathogenic variation (Bruzzone et al, 2002). The genetic cause of the HL wasn’t yet elucidated for this family.
4.3.9. Family OH
This gypsy family is composed of two siblings (fig.12 II:1 and II:2), presenting profound bilateral SNHL phenotype. Their
GJB2 genotype was found to be the same
[p.Trp24X] + [p.Trp24X]. The nonsense
recessive mutation p.Trp24X (fig.7,
supplementary data), localised to the transmembranar domain 1, truncates Cx26
protein, leading to the formation of 24 amino acids, instead of the normal 226 polypeptide. According to literature (Minárik et al, 2003), in individuals homozygous for p.Trp24X, no functional Cx26 channels are present in the cochlea, which has a negative impact on K+ recycling to the endolymph, thus resulting a week or null physiological response to sound stimuli (Minárik et al, 2003). The p.Trp24X mutation is found in high frequency among gypsy populations, namely Spanish Romani, Slovak Romani and Indian, with recent data pointing to the specificity of this mutation to Indian population (Álvarez et al, 2005; Minárik et al, 2003; Padma et al, 2009), since it was found to account for 73% of all pathogenic mutations in GJB2 gene (Mani et al, 2009) in this population. Its higher frequency (2,4%) in Indian population leads to the question if the high carrier frequency of p.Trp24X mutation is also associated with a heterozygote advantage (Mani et al, 2009). Because of its premature STOP codon, p.Trp24X is not expected to form a protein, but a study demonstrate an apparently full-length protein but with defective cellular localisation, being retained in cytoplasm (Mani et al, 2009). This evidence points for the need of performing functional studies even with mutations leading to premature STOP codon. All this data lead us to conclude that the homozygosity of p.Trp24X in both siblings is certainly the cause of their HL phenotype.
4.3.10. Family OI
This family present two daughters (fig.13 II:1 and II:2) with severe and moderate to severe nonsyndromic bilateral SNHL, respectively. The sisters were found to carry two different coding variants, and thus having the same GJB2 genotype [p.Met34Thr] +
[p.Trp24X ] + [p.Trp24X] Unknown genotype Unknown genotype I:1 I:2 II:1 II:2 [p.Trp24X ] + [ p.Trp24X]
[p.Arg184Pro], (Gonçalves et al, 2012, C). Their hearing father (fig.13 I:1) is a p.Arg184Pro carrier, while their hearing mother (fig.12 I:2) is a p.Met34Thr carrier.
Mutation p.Arg184Pro (fig.8, supplementary data) was previously reported as pathogenic and with a
recessive pattern of transmission (Martinez et al, 2009; Bruzzone et al, 2002). This mutation occurs at extracellular loop 2 of Cx26 and prevents not only the traffic of the protein to the membrane but also its correct oligomerization (Martinez et al, 2009; Bruzzone et al, 2002), thus resulting that no gap junction channels are formed.
Because p.Arg184Pro mutation is a recessive one, its presence in only one allele in both sisters does not explain by itself the HL of them. However, its association with p.Met34Thr leads us to assume that the genotype [p.Met34Thr] + [p.Arg184Pro] could be the cause of the deafness observed in both individuals.
The data concerning this family and case NY, above described (section 4.3.6), thus point to a possible pathogenic role of p.Met34Thr variant (discussed in section 4.3.2), either as a recessive allele or as a polymorphism which increases the severity of the phenotype of a recessive monoallelic mutation.
4.3.11. Case OM
This single individual presented profound SNHL was previously found to be heterozygous for c.35delG mutation in section 4.2. Due to this heterozygosity, this proband’s DNA was sequenced for GJB2 gene and his genotype was found to be [c.35delG] + [p.Trp172X]. The p.Trp172X mutation (fig.9, supplementary data) is localised to the extracellular loop 2 of Cx26 protein and leads to a premature STOP codon. This mutation was only reported twice. It was reported for the first time on a Brazilian patient present in an homozygous state (Pfeilsticker et al, 2004) and the second time in a multicentre study from Snoeckx and collaborators (Snoeckx et al, 2005).
Due to the presence of c.35delG mutation that leads to a frameshift causing a premature chain termination and its conjugation with p.Trp172X mutation that causes a premature STOP codon, the genetic cause for the profound HL phenotype presented by this proband can be due to the compound heterozygosity [c.35delG] + [p.Trp172X].
[p.Met34Thr] + [p.Arg184Pro] [=] + [p.Met34Thr] [=] + [p.Arg184Pro] I:1 I:2 II:1 II:2 [p.Met34Thr] + [p.Arg184Pro]
4.3.12. Family PD
This family is composed of an eight year old proband (fig.14 II:1) presenting moderate bilateral SNHL and his normal hearing parents. Proband’s
GJB2 genotype was found to be [p.Val37Ile] +
[p.Asn206Ser], having inherited from the father the mutation p.Val37Ile and from the mother (fig.14 I:2) the p.Asn206Ser mutation.
The p.Val37Ile mutation (fig.10,
supplementary data) is localised to the transmembranar domain 1 of Cx26 protein (Martinez, 2009) and its pathogenic role is still controversial. This mutation was first described as being a polymorphism, because of its high carrier frequency in the general population (Kelley et al, 1998). However, several studies have proposed that p.Val37Ile is in fact pathogenic with a reduced penetrance pattern because it was found overexpressed among individuals with mild to moderate HL (Snoeckx et al, 2005; Pollak et al, 2007; Ma et al, 2010). Pollak also hypothesised that p.Val37Ile mutation can be related to relatively late onset and progression of HL (Pollak et al, 2007).
The p.Asn206Ser mutation (fig.11, supplementary data) is located in transmembranar domain 4 of Cx26 (Martinez et al, 2009), and was associated with a higher proportion of moderate or mild HL (Marlin et al, 2005). This mutation can lead to the formation of functional channels with levels of conductance similar to those observed in wild-type Cx26 channels, but with small differences in gating or permeability (Martinez et al, 2009).
Having into consideration the above data on both mutations, the phenotype presented by this proband may be considered as the result of the genotype [p.Val37Ile] + [ p.Asn206Ser] observed.
4.4. Summing up on GJB2 screening
The results described above, whereupon 25,7% of the studied families (18/70) were found to harbour mutations in GJB2 gene are according to previous determined ones, in which the DFNB1 was a likely cause to explain 20,8% of the HL phenotype in the Portuguese population (Matos, 2012). In the present study, we identify two families with HL due to c.35delG, and 6 other families where HL is due to other genotypes identified in GJB2 gene, accounting for a total of 11,4% of elucidated cases. Figure 15 shows the Cx26 variants found
[p.Val37Ile] + [p.Asn206Ser] [=] + [p.Asn206Ser] [=] + [p.Val37Ile]
I:1 I:2
II:1
during this work, with special emphasis for the novel mutation, p.Leu213X, reported here for the first time.
Sixty genes were mapped until now and more than 100 loci were identified as being responsible for nonsyndromic HL cases (Minami et al, 2012). So, it can always be considered the possibility that the genetic cause of HL for the remaining families could due to mutations in other genes. p.Met34Thr p.Arg184Pro p.Ile140Ser p.Trp24X p.Val37Ile p.Asn206Ser c.333-334delAA p.Leu213X p.Gly160Ser p.Arg127His c.35delG p.Lys224Gln p.Val95Met p.Phe83Leu p.Trp172X
Figure 15 – Schematic representation of Cx26 protein. The arrows indicate the mutations and variants identified in this study and the novel mutation p.Leu213X is shown in red box. The extent of amino acid conservation is colour-coded, with residues shown in blue (1-2) not conserved and rapidly evolving. Residues in white (3-6) show an average degree of conservation and residues in red (7-9) are highly conserved and are slowly evolving. The degree of conservation of the polymorphic residues was analysed using ConSeq, the sequence only variant of Rate4Site, an algorithmic tool for the identification of functional regions in proteins. Adapted from Mani et al, 2009.
